AI for professional services firms: start with how an engagement moves through your firm.
By Chris McCarthy, Founder & Chief AI Officer, OB.1
Your partners want to talk about AI. Someone asks the fair question: how should the firm use it, and where do we start?
Here's where my head is at. AI for professional services firms starts with how an engagement moves through your firm today, from intake to delivery, review and billing. Write down who does each step and which rule they follow. Then rank what to fix first. At OB.1, that first step is an Operational Blueprint.
This post is for accounting, engineering, consulting and similar firms. It's about how the work moves: intake, scoping, staffing, documents, review and billing. It isn't advice on ethics, confidentiality or the rules of your profession. Accounting firms can start with AICPA resources. Engineering firms can start with their state licensing boards. Other firms can start with their own counsel.
The short answer
Start with one process your partners care about. For many firms, the path an engagement takes from intake to billing is a strong first pick. Walk it with the people who run it. Write down each step, who owns it and the rule they follow. Then rank what to fix. Pick AI tools last.
Before you buy AI, write down how your company runs.
I wrote this line for any business, and it fits a professional services firm well.
AI tools run on rules and data. They need to know who sets the scope, who approves the plan and where the files live.
In a lot of firms, those rules live in people's heads. They work fine day to day. A tool can't read them, though.
So the first job is to get the work on paper. Write the steps down. Name the owner of each one. Note the rules they use to make calls.
The rule I work by is simple. Decide how AI will be used before anyone picks a tool. More on that step in how to decide how AI will be used.
Where AI for professional services firms should start
An engagement touches most of the firm. A client asks for help. A partner decides whether to take it on. Someone writes the scope and the proposal. Then a project lead builds the plan and the schedule. Staffing picks the team. A partner reviews the work before it goes out. Billing sends the invoice.
Every handoff in that chain has a rule behind it. Some are written down, and many are not.
When you map that path, you see the whole firm in one process. You see the handoffs, the owners and the gaps. That's also how to use AI in an accounting firm: map how a return or a close moves before you pick a tool.
Before you automate, ask the person who knows.
In many firms, project management specialists keep projects on track. The U.S. Bureau of Labor Statistics lists their tasks. They talk with clients to set project goals. They build plans with schedules and staff, and they assign duties to the team. Source: BLS Occupational Outlook Handbook.
So ask your project lead how a project plan gets built. Ask how they size the work, who they check with on staffing and what they confirm before the plan goes to the client.
They know the steps no form captures. Writing it down with them helps the whole team and the next hire.
The same method in a manufacturing company
We use the same method in manufacturing companies. A worked example shows how it carries over.
Take how a manufacturing company builds a quote. Sales takes the request. Engineering reads the drawing. Purchasing checks material. Scheduling checks capacity. Finance sets the margin.
Swap the names, and the shape looks a lot like a new engagement. A request comes in. A few people size it and check who is free. Someone approves the terms, and the work starts.
In both, the rules often sit with a few people. In both, writing them down comes before picking a tool. For the full walk-through, read how your company builds a quote. For the method from the top, read where a manufacturing company should start with AI.
How to start, one process at a time
Pick one process your partners care about
Client intake and scoping is a strong default. Document collection and partner review are good picks too. Choose the one that takes the most partner time.
Walk it with the people who run it
Sit with the people who do the work. Ask them to show you a recent project, start to finish.
Write down each step, who owns it and the rule they follow
One line per step. Who does it, what they decide and how they decide it.
Note where the data lives
Some of it sits in your project or time system. Some is in email, shared drives, spreadsheets or one person's notes. Write down where each piece sits.
Rank what to fix first
Look for steps that wait on one person, rules nobody wrote down and data your team types twice. Rank them by how much they slow the firm.
Then decide where AI is the right fix
Some fixes need a clear owner or a written rule. Some need a better template. A few will be good work for AI. We'll tell you when AI is the wrong fix.
For a closer look at the mapping itself, read how to map how work moves.
What you'll know when you're done
You'll know how work moves through your company and what to fix first.
Your partners get a shared picture, and the firm gets a clear first target. Any AI tool you pick later has something steady to work from.
Want a gut check first? Read how to tell if you're ready for AI.
How OB.1 helps
We start by looking at how your firm runs today.
The first step is an Operational Blueprint. It starts with a 30-minute Blueprint Session. It gives you a diagnosis of your platforms, process and people, and a plan your team can act on.
We map how work moves through your company and who owns each rule, then rank what to fix first.
No rush on any of this. Whenever you're ready, let's talk.
Questions professional services leaders ask
How do consulting firms use AI?
The best place to start is the work. Map how an engagement moves from intake to billing. Write down each step, who owns it and the rule it follows. Then rank what to fix first. Choose tools after that.
What is the best AI tool for professional work?
Firms use a range of tools, and the right one depends on the work. Start with the work, then pick the tool. Once the steps and rules are on paper, you can judge any tool against them.
Who should be in the room when we map the work?
The people who do it. In many firms, that means project leads, the partner who signs off on scope and the people who handle billing.
Does this work for accounting, civil engineering and consulting firms?
It can. Each one scopes work, staffs it, reviews it and bills for it. The steps differ, and the method stays the same.
Does our accounting firm need an AI policy?
That's a question for your own leaders and advisors. For accounting firms, AICPA's small firm AI policy template is a good place to start. Engineering firms can check with their state licensing boards, and other firms with their own counsel. A map of how work moves gives any policy something real to point to.
How does a firm compare to a manufacturing company here?
The method is the same. A manufacturing company maps how a quote gets built. A firm maps how an engagement moves from intake to billing. Both write down the rules before they choose tools.